Late last week, the White House confirmed that Amazon, Anthropic, Google, Inflection, Meta, Microsoft, and OpenAI have agreed to self-regulate their AI advancements. These seven companies have signed a document called Ensuring Safe, Secure, and Trustworthy AI, which is built on three pillars: safety, security, and trust.
Generative AI has become extremely popular in the past nine months. According to the Summer 2023 Fortune/Deloitte CEO Survey, 37% of CEOs from various industries are already implementing generative AI to some extent, while 55% are evaluating and experimenting with it. Many CEOs believe that generative AI will enhance efficiency (79%) and increase growth opportunities (52%). However, AI is a two-edged sword due to security and privacy concerns, as well as its potential for mis or disinformation.
To address these concerns, the signed document focuses on three key areas: safety, security, and trust.
How Will Companies Self-Regulate AI Development?
To ensure safety, the companies have committed to red-teaming efforts, which involve identifying and eliminating societal risks and national security concerns. This includes assessing the tech’s applicability to developing biological, chemical, and radiological weapons, as well as addressing cybersecurity risks, bias or discrimination, and the risk of self-replication.
Additionally, the signatories are dedicated to information-sharing efforts among themselves and with the government. They will establish a forum or mechanism to develop, advance, and adopt shared standards and best practices for AI safety.
Under the security commitments, the companies have agreed to establish both external and internal threat detection programs. They also plan to incentivize third-party vulnerability detection and responsible disclosure through bug bounty programs, contests, or prizes.
To ensure trust, the companies will implement provenance and/or watermarking systems for any audio or visual content generated by their AI tools. The document requires regular safety evaluations that explain the capabilities, limitations, and appropriate use of all AI service versions.
Furthermore, the signatories will support research and development initiatives to overcome major societal challenges, including climate change, early cancer detection and prevention, and combatting cyber threats.
It’s important to note that this document applies only to generative AI tools and models that are more advanced than current industry standards. This includes GPT-4, Claude 2, PaLM 2, Titan, and DALL-E 2 (for image generation).
Representatives of these seven companies have expressed their commitment to this initiative on their respective blogs. Meta’s president of global affairs, Nick Clegg, stated, “Meta welcomes this White House-led process, and we are pleased to make these voluntary commitments alongside others in the sector. They are an important first step in ensuring responsible guardrails are established for AI, and they create a model for other governments to follow.”
Status of AI Regulations in Other Countries
While the United States is in the process of shaping federal AI regulations, the European Union (EU) is on track to pass its AI Act later this year. In July, China released updated guidelines to regulate AI, which will take effect on August 15, 2023.
The White House stated that they will work with allies and partners to establish a strong international framework for governing AI development and use. They have already consulted on the voluntary commitments with several countries, including Australia, Canada, France, Germany, Japan, and the UK.
However, there are concerns about China gaining a “head start” in AI self-regulation. Sen. Mark Warner expressed his worries that China is ahead of the game in terms of self-regulating AI within their nation-state. He also mentioned that China has specific legislation on AI, which is even more advanced than Europe’s regulations.
Should Companies Self-Regulate AI Developments?
The question of whether companies should be allowed to regulate their AI developments is open for debate. What are your thoughts on this matter? Share your opinions with us on LinkedIn, Twitter, or Facebook. We would love to hear from you!
It’s encouraging to see major tech companies taking the initiative to self-regulate their AI developments. This demonstrates their commitment to ensuring the safe and responsible use of AI technology. By voluntarily signing the Ensuring Safe, Secure, and Trustworthy AI document, these companies are setting a positive example for other industry players and governments worldwide.
At GPT News Room, we strive to provide the latest updates and insights on AI developments and their impact on society. Stay tuned for more news and articles about the future of AI by visiting our website at GPT News Room.
[Link to GPT News Room: https://gptnewsroom.com]